Cooperative learning with joint state value approximation for multi-agent systems
This paper relieves the 'curse of dimensionality' problem, which becomes intractable when scaling rein- forcement learning to multi-agent systems. This problem is aggravated exponentially as the number of agents increases, resulting in large memory requirement and slowness in learning speed. For coo...
Saved in:
Published in | Journal of control theory and applications Vol. 11; no. 2; pp. 149 - 155 |
---|---|
Main Authors | , , , |
Format | Journal Article |
Language | English |
Published |
Heidelberg
South China University of Technology and Academy of Mathematics and Systems Science, CAS
01.05.2013
School of Information Science and Engineering, Central South University, Changsha Hunan 410083, China |
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | This paper relieves the 'curse of dimensionality' problem, which becomes intractable when scaling rein- forcement learning to multi-agent systems. This problem is aggravated exponentially as the number of agents increases, resulting in large memory requirement and slowness in learning speed. For cooperative systems which widely exist in multi-agent systems, this paper proposes a new multi-agent Q-learning algorithm based on decomposing the joint state and joint action learning into two learning processes, which are learning individual action and the maximum value of the joint state approximately. The latter process considers others' actions to insure that the joint action is optimal and supports the updating of the former one. The simulation results illustrate that the proposed algorithm can learn the optimal joint behavior with smaller memory and faster leamin~ soeed comoared with friend-O learnin~ and indet~endent learning. |
---|---|
Bibliography: | 44-1600/TP This paper relieves the 'curse of dimensionality' problem, which becomes intractable when scaling rein- forcement learning to multi-agent systems. This problem is aggravated exponentially as the number of agents increases, resulting in large memory requirement and slowness in learning speed. For cooperative systems which widely exist in multi-agent systems, this paper proposes a new multi-agent Q-learning algorithm based on decomposing the joint state and joint action learning into two learning processes, which are learning individual action and the maximum value of the joint state approximately. The latter process considers others' actions to insure that the joint action is optimal and supports the updating of the former one. The simulation results illustrate that the proposed algorithm can learn the optimal joint behavior with smaller memory and faster leamin~ soeed comoared with friend-O learnin~ and indet~endent learning. Multi-agent system; Q-learning; Cooperative system; Curse of dimensionality; Decomposition ObjectType-Article-2 SourceType-Scholarly Journals-1 ObjectType-Feature-1 content type line 23 |
ISSN: | 1672-6340 1993-0623 |
DOI: | 10.1007/s11768-013-1141-z |